Doubt regarding abuse-related appraisals and identification with the aggressor as predictors of complex PTSD in female child abuse survivors

ABSTRACT

Background: Child abuse (CA) is a significant risk factor for trauma-related psychopathology, with potential outcomes that extend beyond posttraumatic stress disorder (PTSD) to include complex PTSD (CPTSD) – a condition characterized by disturbances in self-organization (DSO). This trauma can also lead to identification with the aggressor (IWA), where survivors internalize the perpetrator’s beliefs, perspectives, and behaviors, as well as doubt regarding abuse-related appraisals (DARA), which reflects uncertainty in interpreting aspects of the abuse. Although IWA and DARA have been proposed as potential contributors to trauma-related symptomatology, their predictive roles have not been empirically examined.

Objective: This two-wave study explored the implications of IWA and DARA for subsequent PTSD and DSO symptoms.

Method: The current study was conducted among 273 adult female CA survivors, aged 18–53 (M = 33.01, SD = 9.78). Participants completed online self-report measures assessing IWA and DARA at the first measurement (T1) and PTSD and DSO symptoms at two time points (T1 and T2).

Results: The results revealed positive associations between IWA and DARA at T1 and PTSD and DSO symptoms at T2. Analyses further indicated that the IWA component, which involves the replacement of one’s agency with that of the perpetrator at T1, predicted variance in PTSD and DSO at T2 (ES = 0.15 and 0.15, respectively). Additionally, the DARA component, which reflects doubt regarding the abuse at T1, predicted variance in DSO symptoms at T2 (ES = 0.17). These effects remained significant even after accounting for polyvictimization, PTSD, and DSO at T1.

Conclusions: IWA and DARA may be important psychological factors contributing to survivors’ vulnerability to trauma-related psychopathology.

HIGHLIGHTS

• Identification with the aggressor relates to PTSD and DSO symptoms

• Doubt regarding abuse-related appraisals relates to PTSD and DSO symptoms

• Replacing one's agency with that of the perpetrator predicts PTSD and DSO symptoms

• Doubt regarding the abuse predicts DSO symptoms

Child abuse (CA) – including physical, emotional, and sexual maltreatment – is a pervasive global issue that significantly compromises the well-being of millions of children worldwide (Stoltenborgh et al., 2015). This form of early trauma can severely disrupt developmental processes and is associated with a range of long-term health consequences that often extend into adulthood (Sahle et al., 2022; Wegman & Stetler, 2009). The enduring mental health effects of CA are extensive, including increased risks for depression, anxiety, dissociation, substance use disorders, eating disorders, nonsuicidal self-injury, sexual dysfunction, suicidality, and trauma-related conditions (Angelakis et al., 2019; Ford & Gómez, 2015; Gewirtz-Meydan & Lahav, 2020a, 2020b; Lindert et al., 2014; Messman-Moore & Bhuptani, 2017; Nelson et al., 2017; Vonderlin et al., 2018; Wang et al., 2023).

Survivors of CA are at heightened risk of developing Posttraumatic Stress Disorder (PTSD), characterized by three core symptom clusters directly related to the traumatic experience: re-experiencing the trauma in the present, avoidance of trauma-related cues, and a persistent sense of heightened threat (The International Classification of Diseases, 11th edition, 2018). However, when trauma is prolonged, repeated, and interpersonal – as is often the case with CA – it can result in pervasive and chronic disturbances in self-organization that extend beyond the scope of PTSD, and are reflected in Complex PTSD (CPTSD; Cloitre, 2020; Herman, 1992; Maercker et al., 2022). To capture these broader impacts, the ICD-11 introduced Complex PTSD (CPTSD), a diagnosis that encompasses the three PTSD symptom clusters and three additional clusters reflecting disturbances in self-organization (DSO): affect dysregulation (e.g. intense emotional reactivity, outbursts of anger, reckless or self-harming behavior), negative self-concept (e.g. feelings of worthlessness or failure), and difficulties in forming and maintaining relationships (e.g. avoidance of relationships or intense, unstable interpersonal connections).

The conceptualization of complex trauma and its sequelae has been shaped by several foundational contributors in the field. Herman’s (1992) seminal work laid the groundwork for understanding the pervasive effects of prolonged interpersonal trauma and introduced the early diagnostic construct of Disorders of Extreme Stress Not Otherwise Specified (DESNOS), emphasizing disturbances in affect regulation, identity, and interpersonal functioning. Building on this foundation, the ICD-11 formulation of Complex PTSD was developed through the contributions of Cloitre, Maercker, and colleagues, who delineated CPTSD as a diagnosis distinct from PTSD and highlighted the central role of disturbances in self-organization (Cloitre, 2020; Cloitre et al., 2013; Maercker et al., 2013). Large-scale investigations across diverse populations further advanced the field by providing robust empirical support for the differentiation between PTSD and CPTSD (e.g. Cloitre et al., 2019; Maercker et al., 2018).

Despite growing consensus, several controversies continue to characterize the field. A central debate concerns whether CPTSD constitutes a distinct diagnostic entity or represents a severe subtype of PTSD. This debate is reflected in the divergence between ICD-11 and DSM-5: whereas ICD-11 recognizes CPTSD as a separate diagnosis alongside PTSD, DSM-5 does not include CPTSD as a distinct category and instead conceptualizes trauma-related disturbances within the broader PTSD framework or related disorders (Brewin et al., 2017; Resick et al., 2012). Additional controversies pertain to the boundaries between CPTSD and borderline personality disorder, given overlapping features in affect regulation and interpersonal instability (e.g. Ford & Courtois, 2014; Maercker et al., 2022), as well as questions about the universality of the CPTSD construct across cultures. Some studies indicate variability in symptom structure and presentation (Ben-Ezra et al., 2018; Kazlauskas et al., 2018). These debates underscore the importance of continued empirical work aimed at clarifying the psychological processes underlying CPTSD.

Studies among non-clinical adult community samples have reported prevalence rates of CPTSD ranging from 0.5% to 7.7% (Ben-Ezra et al., 2018; Cloitre et al., 2019; Hyland et al., 2021; Hyland, Karatzias et al., 2020; Hyland, Shevlin et al., 2020; Maercker et al., 2018). In contrast, significantly higher rates – up to 36% – have been observed among adults receiving mental health treatment (Møller et al., 2020). Moreover, evidence suggests that CPTSD is more commonly associated with childhood trauma than PTSD (Cloitre et al., 2019; Hyland et al., 2017; Leiva-Bianchi et al., 2023).

Given that not all individuals who experience CA develop CPTSD, and that symptom severity varies considerably among those who do, identifying relevant risk factors is crucial. Previous research has highlighted several variables that may increase vulnerability to CPTSD following CA, including the severity and chronicity of the abuse (Brewin et al., 2017; Cloitre, 2020; Karatzias et al., 2022). However, the specific psychological processes through which CA contributes to the development of CPTSD remain insufficiently understood. The current two-wave study aims to bridge this knowledge gap by exploring the effects of two phenomena–identification with the aggressor (IWA; Lahav, Talmon, and Ginzburg, 2019a) and doubt regarding abuse-related appraisals (DARA; Lahav, Huberman, et al., 2025) – in predicting symptoms of CPTSD, specifically both PTSD and DSO symptoms, following CA.

The concept of Identification with the Aggressor (IWA), originally introduced by Ferenczi (1932, 1933), has been proposed as a survival mechanism that enables victims to endure abuse by internalizing and adopting the perpetrator’s experience (Coates & Moore, 1997; Frankel, 2002; Lahav, Talmon, & Ginzburg, 2019a). This psychological mechanism helps victims preserve a sense of connection with the perpetrator and may emerge in contexts marked by significant power imbalances – such as when escape or resistance is impossible (Frankel, 2002, 2018).

IWA comprises four interrelated components: replacing one’s agency with that of the perpetrator; becoming hypersensitive to the perpetrator; adopting the perpetrator’s experience concerning the abuse; and identifying with the perpetrator’s aggression (Lahav, Talmon, & Ginzburg, 2019a). As a means of anticipating and surviving the abuse, victims may develop dissociated self-states that identify with the aggressor (Lahav, Talmon, & Ginzburg, 2019a; Lahav, Talmon, Ginzburg, and Spiegel, 2019b). Abuse victims experience the replacement of their own agency with that of the perpetrator, resulting in a disconnection from their feelings, urges, and needs, leading them to become passive and obedient in both behavior and thought. They may develop heightened sensitivity towards their perpetrators, becoming hyper-attuned to their abuser’s moods, expectations, and preferences, striving to appease them. In adopting the perpetrator’s experience of the abuse, victims may internalize their abuser’s justifications – minimizing, denying, or rationalizing the harm inflicted while blaming themselves (Lahav, 2023; Lahav et al., 2017; Siegel et al., 2024; Sultana & Lahav, 2023). A further dimension involves identification with the perpetrator’s aggression, leading victims to turn this aggression inward or outward (Lahav, 2021b; Lahav et al., 2020).

While IWA is considered an automatic response that facilitates survival during abuse (Frankel, 2002), it can become ingrained and persist long after the abuse has ceased – at which point it may become profoundly harmful (Lahav, Talmon, & Ginzburg, 2019a). Because IWA involves the splitting off of aspects of one’s experience (i.e. dissociation), it has been argued to impede trauma processing (Frankel, 2002). Additionally, forming emotional ties with the perpetrator and internalizing their perspective and emotions regarding the abuse has been suggested to lead to re-experiencing the trauma, thereby hindering recovery and intensifying psychological distress (Lahav, 2021a; Lahav, Talmon, & Ginzburg, 2019a). Empirical findings support these assertions, showing that IWA is associated with a range of adverse outcomes in adult survivors of CA – including dissociation, non-suicidal self-injury, sexual revictimization, and suicidal ideation and behavior (Lahav, 2021b; Lahav, Talmon, & Ginzburg, 2019a; Lahav, Talmon, Ginzburg, & Spiegel, 2019b; Lahav et al., 2020; Rosenberg et al., 2023). Moreover, a recent study demonstrated a relationship between IWA and both PTSD and DSO symptoms, suggesting that the replacement of one's agency with that of the perpetrator as part of IWA serves as a risk factor for both PTSD and CPTSD classifications (Lahav, Cloitre, et al., 2025). However, as existing research has predominantly employed cross-sectional designs, the question of whether IWA predicts PTSD and CPTSD symptomatology remains unresolved.

Another distinct yet related phenomenon that may also contribute to the development of CPTSD is Doubt Regarding Abuse-Related Appraisals (DARA; Lahav et al., forthcoming; Lahav, Huberman, et al., 2025). DARA is a novel theoretical framework developed by the first author. Conceptualized as a potential outcome of abusive experiences such as CA, it denotes survivors’ uncertainty in appraising key aspects of the abuse. It is important to emphasize that DARA does not refer to questioning whether the abuse occurred (Porat-Moeller et al., 2025), but rather to uncertainty surrounding one’s appraisals of key aspects of the abusive experience. DARA is conceptually distinct from other maladaptive appraisals commonly observed among trauma survivors, such as negative self-cognitions (Foa et al., 1999; Gómez de La Cuesta et al., 2019; Greenblatt-Kimron et al., 2023; Serier et al., 2023; Shahar et al., 2013) or trauma appraisals (Kucharska, 2017; Ogle et al., 2016). Unlike these appraisals, which reflect the content or valence of one’s beliefs, DARA reflects difficulty in forming or maintaining stable evaluations of central aspects of the abuse. Thus, while maladaptive appraisals are characterized by their negative tone, DARA is defined by a lack of certainty regarding abuse-related judgments.

Specifically, DARA encompasses survivors’ uncertainty regarding three key aspects of the abuse: the nature of the abuse itself (e.g. its severity, legitimacy, or how it should be labeled), the self during the abuse (e.g. one’s sense of control or ability to prevent it), and the perpetrator (Lahav, Huberman, et al., 2025; e.g. whether they are dangerous or trustworthy). According to DARA theory, several processes inherent in abusive experiences, such as CA, may contribute to the development of DARA. The overwhelming nature of trauma, coupled with dissociative reactions during the abuse (e.g. derealization and depersonalization), can disrupt cognitive processing (Brewin et al., 1996; Ehlers & Clark, 2000; Kolk & Fisler, 1995), potentially resulting in a breakdown of the victim’s ability to process, integrate, and categorize the experience coherently (van der Kolk, 2005). Additionally, the relational dynamics between the abused child and the perpetrator – often marked by asymmetrical power distribution, emotional complexity, elusiveness, and unpredictability – may further contribute to the development of DARA. Perpetrators frequently undermine victims’ perceptions and deny the occurrence of abuse (Davies, 2019; Harsey et al., 2017). Moreover, they may label abusive behavior in distorted ways, such as referring to brutality as affection and exploitation as a sign that the child is special (Jackson et al., 2015), thereby compromising the victim’s capacity to trust their own experiences. In addition, the victim’s identification with the aggressor as a way to survive the abuse may further fuel DARA, as it coexists in a disconnected manner alongside other self-states that hold contradictory feelings and views regarding the abuse. Thus, whereas the self-state that identifies with the aggressor may normalize the abuse, blame the child for the abuse, and view the perpetrator as righteous, other self-states may appraise the situation differently. Consequently, children who experience abuse may develop marked confusion concerning the nature of the abuse, their self during the abuse, and their understanding of the perpetrators, rendering these fundamental aspects subject to considerable doubt.

Given the chronic and cumulative nature of these processes, DARA is posited to persist well beyond the period of abuse (Lahav, Huberman, et al., 2025). As a result, survivors may continue to experience pervasive doubt – not only regarding the abuse itself but also concerning their responses during the abuse and their perceptions of the perpetrator and themselves – even many years after the traumatic events have ended. According to DARA theory, this sense of uncertainty fosters confusion and misperceptions and heightens posttraumatic distress among survivors. In particular, difficulties in trusting their own judgment regarding the abuse may hinder survivors from constructing a coherent narrative about the traumatic events and their aftermath. This can leave them in a persistent state of confusion, ambiguity, and disorientation, impeding trauma processing and potentially heightening their vulnerability to trauma-related symptoms (Lahav et al., forthcoming; Lahav, Huberman, et al., 2025). Findings from a recent study on DARA among childhood abuse survivors have shown that DARA can persist for years after the abuse has ended. The study also identified associations between higher levels of DARA and increased distress, including feelings of guilt, shame, dissociation, and symptoms of PTSD among survivors (Lahav, Huberman, et al., 2025).

Nevertheless, to the best of our knowledge, the relationship between DARA and CPTSD has yet to be investigated. Furthermore, no empirical investigation has uncovered the contribution of both IWA and DARA in predicting subsequent PTSD and DSO symptoms among CA survivors. Given that both IWA and DARA represent enduring consequences of abuse that often become deeply ingrained in survivors’ psychological frameworks, they may not only intensify PTSD symptoms but also adversely impact survivors’ sense of self and interpersonal functioning – core features reflected in the DSO clusters of CPTSD. Therefore, examining the predictive role of IWA and DARA in relation to both PTSD and DSO symptoms is of significant importance. To bridge this knowledge gap, the current two-wave study, which includes two time points (T1 and T2), had two primary objectives: (1) to examine the relationships between IWA and DARA and subsequent PTSD and DSO symptoms; and (2) to evaluate the predictive effects of IWA and DARA at T1 on subsequent (T2) PTSD and DSO symptoms, above and beyond the effects of polyvictimization and T1 PTSD and DSO symptoms.

2. Methods

2.1. Participants and procedure

The current research is part of a larger multi-wave study examining the implications of CA on adults’ mental health. This study utilized a convenience sample of Israeli adults aged 18 years and older residing in Israel, using an online survey. Data for the current study were collected at two time points: from December 2022 to March 2023 (T1) and from April 2023 to August 2023 (T2). The Tel Aviv University institutional review board (IRB) approved all procedures and instruments.

The survey was administered via Qualtrics software (QualtricsLabs, Inc., Provo, UT, US), and was advertised as a study exploring the consequences of challenging childhood experiences on adult health. Participants accessed an information page outlining the study's purpose, types of questions, and a consent form. Informed consent was obtained electronically, and participants were informed they could withdraw at any time. They were also provided with contact details for the research team and a list of Israeli organizations offering support for survivors of childhood abuse. Recruitment primarily occurred via social media (Facebook), with a small portion (∼10%) of participants recruited through the Sona system of Tel Aviv University, an academic platform for university student research participation. Eligibility was limited to individuals aged 18 and older residing in Israel. The Facebook advertisement included a headline, main text, and a survey link. Social media platforms like Facebook are increasingly utilized in mental health research to reach hard-to-access populations, including survivors of childhood abuse, who may not engage with traditional recruitment channels. Given that the vast majority of the sample was recruited via Facebook, the current sample primarily reflects community-based participants rather than a student population.

To ensure data validity and authenticity, multiple quality assurance procedures were applied. First, the ‘Prevent Ballot Box Stuffing’ feature in Qualtrics was enabled to prevent multiple submissions from the same device. Additionally, 12 duplicate cases were identified and removed based on matching IP addresses. Standard data quality checks were implemented to identify inattentive, careless, or non-genuine responses. These included: (1) Exclusion of implausible or out-of-range values in quantitative items (e.g. age, scale responses); (2) Removal of nonsensical or irrelevant content in open-text fields (e.g. random characters, unrelated content); (3) Enforcement of a minimum completion time threshold to detect inattentive responses. To further enhance data authenticity, only participants who provided a valid email address at T1 and responded to the personalized follow-up invitation at T2 were included in the final two-wave sample. This two-stage participation process served as a basic form of identity verification and reduced the risk of bot-generated or fraudulent responses.

At the end of the T1 survey, participants were given the option to leave their email addresses if they were willing to participate in the second wave (T2). To protect privacy, email addresses were stored separately from response data in a secure, password-protected file. Participants who provided email addresses were recontacted at T2 and invited to complete a second survey. The average survey duration was 30 min at both time points. As compensation, participants could enter a draw for one of ten 200 NIS gift vouchers or, if recruited via Sona, receive academic credit.

At T1, a total of 1,868 individuals accessed the survey. Of these, 1,209 completed only part of the questionnaires and were excluded. An additional 12 duplicate entries (identified via matching IP addresses) were also removed. Among the remaining participants, 776 (64.1%) provided an email address for T2 follow-up and were subsequently invited to participate. Of these, 495 accessed the T2 survey (63.8% response rate), and 456 completed it (58.8%).

Among the full responders, 296 participants met the criteria for a history of CA, based on the Childhood Trauma Questionnaire-Short Form (CTQ-SF; Bernstein et al., 2003), and provided complete data on the study variables. The majority identified as women (92.2%), with 5.7% identifying as male and 2.0% as another gender. Due to the small number of non-female respondents and to avoid statistical bias due to gender imbalance, only female participants were included in the current analysis. The final sample consisted of 273 adult women with a history of CA who completed both waves of the study and all relevant measures.

Of this final sample, 47.3% reported a history of childhood physical abuse, 64.5% reported childhood sexual abuse, and 91.6% reported emotional abuse. Most participants (75.5%) experienced polyvictimization (two or more types of abuse), and 89.4% reported recurrent abuse. The average severity of abuse was 35.81 (SD = 11.88). As shown in Table 1, most participants were secular, employed part- or full-time, had below-average income, and were not in a romantic relationship.

Table 1.

Description of demographic characteristics among participants (n = 273).

	M (SD) or n (%)
Age, M (SD)	33.01 (9.78)
Education (years), M (SD)	14.58 (3.43)
Relationship status, n (%)
In a relationship	123 (45.0)
Not in a relationship	143 (52.4)
Other	7 (2.6)
Religiosity, n (%)
Secular	206 (75.5)
Religious/traditional	49 (17.9)
Other	18 (6.6)
Employment status, n (%)
Working in a full or part-time job	183 (67.0)
Not working	50 (18.3)
Other	40 (14.7)
Income, n (%)
Below-average income	157 (57.5)
Average income or above	116 (42.5)
Recurrence of the abuse
Recurrent abuse	244 (89.4)
Single episode of abuse	29 (10.6)
Polyvictimization
Single type of abuse	67 (24.5)
Two or more types of abuse	206 (85.5)

2.2. Measures

2.2.1. Background variables (T1)

Participants completed a brief demographic questionnaire assessing age, education, and relationship status.

2.2.2. Child abuse and polyvictimization (T1)

History of CA was assessed using the physical, sexual, and emotional abuse subscales of the CTQ-SF (Bernstein et al., 2003). Participants rated the frequency of these experiences during childhood on a 5-point Likert scale (1 = never true to 5 = very often true). Classification of CA was based on cutoff scores: ≥9 for emotional abuse, ≥8 for physical abuse, and ≥6 for sexual abuse (Tietjen et al., 2010). Participants exceeding any of these cutoffs were classified as having a history of CA.

The CTQ-SF demonstrates strong convergent and discriminant validity and high internal consistency (Bernstein et al., 2003). In this study, the scale scores showed excellent internal consistency (α = .90). Additionally, exposure to one type of abuse versus polyvictimization was determined based on participants’ responses to the CTQ-SF.

2.2.3. Identification with the aggressor (T1)

IWA was assessed at T1 using the Identification with the Aggressor Scale (IAS), a 23-item self-report questionnaire (Lahav, Talmon, & Ginzburg, 2019a). The items were presented to respondents as reflecting ‘possible reactions that people may experience as a result of abuse.’ Participants rated the frequency with which they experienced each manifestation of IWA in relation to their perpetrator on an 11-point Likert-type scale, ranging from 0% (never) to 100% (all the time).

In addition to the total score, calculated by averaging all items, mean scores were computed for four subscales: adopting the perpetrator’s experience concerning the abuse (9 items; e.g. ‘Some people feel that the point of view of their perpetrator is the right one’), identifying with the perpetrator’s aggression (5 items; e.g. ‘Some people feel that they behave as aggressively as their perpetrator’), replacing one’s agency with that of the perpetrator (5 items; e.g. ‘Some people do not know what they want in the presence of their perpetrator’), and becoming hypersensitive to the perpetrator (4 items; e.g. ‘Some people “read the thoughts” of their perpetrator’). The IAS demonstrates high construct and criterion validity, and high internal reliability (Lahav, Talmon, & Ginzburg, 2019a). In this study, internal consistency values were 0.89, 0.94, 0.77, 0.87, and 0.94 for the subscales and the total score, respectively.

2.2.4. Doubt regarding abuse-related appraisals (T1)

Levels of DARA were assessed using The Abuse Doubt Scale (ADS), a 10-item self-report questionnaire (Lahav et al., forthcoming; Lahav, Huberman, et al., 2025). The items were presented to respondents as reflecting ‘doubts that may arise in regard to CA.’ To avoid social desirability bias and ensure that the items assessed the experience of doubt rather than the strength of appraisals, items were phrased to include both possible responses to unanswered questions regarding CA (e.g. I have doubts about whether the offender is a person who is safe or unsafe to be around). Participants were asked to rate, on a 5-point Likert-type scale ranging from 0 (not at all) to 4 (very much), the extent to which they experience each manifestation of DARA in relation to past CA.

In addition to the total score, calculated by averaging all items, mean scores were computed for the three subscales: (1) doubt regarding the abuse, representing uncertainty around appraisals of features or qualities of the abuse (4 items; e.g. ‘I have doubts as to whether or not the offense was severe’); (2) doubt regarding oneself, representing uncertainty about the victim’s/survivor’s own experiences and behavior in the context of the abuse (3 items; e.g. ‘I have doubts as to whether or not I was helpless at the time of the offense’); (3) doubt regarding the perpetrator which represents uncertainty about appraisals of the perpetrator’s characteristic (3 items e.g. ‘I have doubts as to whether the offender is a person who can or cannot be trusted’). The ADS demonstrates high validity and internal reliability (Lahav, Huberman et al., 2025). In this study, internal consistency was 0.90, 0.86, 0.86, and 0.89 for the subscales and total score, respectively.

2.2.5. Symptoms of PTSD and CPTSD (T2)

In this study, we utilized the International Trauma Questionnaire (Cloitre et al., 2018) to assess symptoms of PTSD and CPTSD at T2. This measure is designed to encompass all elements of the ICD-11 diagnostic criteria for both conditions. The ITQ is a self-report scale; respondents are first asked to identify their most distressing traumatic event and indicate how long ago this event occurred. They are then instructed to answer all subsequent questions in relation to this event.

The questionnaire includes six items measuring PTSD symptoms across the clusters of ‘Re-experiencing in the Here and Now,’ ‘Avoidance,’ and ‘Sense of Threat.’ Respondents indicate how much they have been bothered by each symptom in the past month. Additionally, three questions assess functional impairment in social, occupational, and other important life domains affected by these symptoms.

There are also six items measuring DSO symptoms across the clusters of ‘Affective Dysregulation,’ ‘Negative Self-Concept,’ and ‘Disturbed Relationships.’ These items assess how respondents typically feel, think about themselves, and relate to others. Three further items evaluate functional impairment associated with DSO symptoms. All items are rated on a five-point Likert scale ranging from 0 (Not at all) to 4 (Extremely), with a symptom considered present at a score of ≥2 (Moderately) on the scale. In this study, the internal consistency was high: 0.88 for PTSD symptoms and 0.90 for DSO symptoms.

To meet the diagnostic criteria for PTSD or CPTSD, an individual must be trauma-exposed. For a diagnosis of PTSD, at least one symptom must be present from each PTSD cluster, along with at least one indicator of functional impairment associated with these symptoms. To meet the diagnostic criteria for CPTSD, at least one symptom must be present from all six symptom clusters, and functional impairment associated with both PTSD and DSO symptoms must be endorsed. According to the ICD-11 diagnostic guidelines, an individual can receive a diagnosis of either PTSD or CPTSD, but not both. If a person meets the criteria for CPTSD, they do not also receive a diagnosis of PTSD.

2.3. Analytic strategy

The analyses were performed using R software. The total sample size was 273 participants, with no missing data; thus, no imputation procedures were necessary for our study. PTSD and DSO symptoms were assessed at both T1 and T2 to establish baseline symptom severity and adjust for prior levels of posttraumatic distress. This approach allowed for an examination of whether IWA and DARA assessed at T1 were associated with elevated PTSD and DSO symptoms at T2, above and beyond participants’ symptom levels at T1. To explore the relationships between IWA and DARA at T1 and PTSD and DSO symptoms at T1 and T2, we conducted correlation analyses using Pearson's correlation coefficient where applicable, and Kendall's ${\tau }_B$and ${\tau }_C$ for variables limited to fewer than four values (i.e. in the PTSD and DSO subscales).

To examine the contributions of IWA and DARA at T1 in explaining PTSD and DSO symptoms at T2 – beyond the effects of polyvictimization and baseline symptoms – we employed a structural equation modeling (SEM) framework. PTSD and DSO at T2 were modeled as latent variables, each defined by their respective ITQ symptom cluster indicators. This approach allowed us to account for measurement error and estimate the measurement and structural components simultaneously. We assessed key model assumptions before interpreting the results. The multivariate normality of the indicators was tested using Mardia’s test, which indicated non-normality; therefore, we estimated the model with the robust MLM estimator. Multicollinearity among predictors was evaluated using variance inflation factors (all VIF <5). Linearity was inspected through scatterplots of composite factor proxies, and no deviations were detected. Heteroskedasticity was assessed with the Breusch–Pagan test, which indicated homoscedastic residuals. These diagnostics support the stability and validity of the model estimates. The analysis was conducted hierarchically across three models: Model 1 included only IWA and DARA predictors; Model 2 added polyvictimization; and Model 3 further introduced baseline PTSD and DSO symptoms at T1. This stepwise approach allowed us to assess how the inclusion of additional covariates, particularly baseline symptoms, influenced the magnitude and significance of the IWA and DARA structural paths, and whether these predictors retained explanatory power after accounting for variables that could absorb variance in T2 outcomes. Model fit was evaluated using the normed chi-square value, RMSR, RMSEA, and appropriate incremental fit indices (CFI and TLI).

3. Results

3.1. PTSD and CPTSD classification

At T2, 6.2% (n = 17) of the sample were classified with PTSD, while 22.7% (n = 62) were classified with CPTSD. The frequencies of symptom cluster endorsement were as follows: the most commonly endorsed symptom cluster was Affective Dysregulation (67.0%), followed by Disturbed Relationships (59.7%), Sense of Current Threat (58.6%), Avoidance (52.4%), Negative Self-Concept (50.5%), and Re-experiencing (37.4%).

3.2. IWA, DARA and symptoms of PTSD and DSO

Analyses indicated significant correlations between the IWA total score at T1 and PTSD total scores at T1 (r = 0.16, $p<.01),\mathrm{as}\mathrm{well}\mathrm{as}$ the DSO total score at T1 (r = 0.27, $p<.001)$ and T2 (r = 0.18, $p<.01).$The correlation between the IWA total score at T1 and PTSD total scores at T2 was non-significant (r = 0.11, $p=.07)$. Additionally, analyses indicated significant correlations between the DARA total score at T1 and PTSD total scores at T1 (r = 0.19, $p<.01)$and T2 (r = 0.17, $p<.01),\mathrm{as}\mathrm{well}\mathrm{as}$ the DSO total score at T1 (r = 0.29, $p<.001)$ and T2 (r = 0.24, $p<.001$).

Table 2 presents the correlations between components of IWA and DARA and various PTSD and DSO symptoms. As shown, replacing one's agency with that of the perpetrator (IWA) was associated with higher levels of all PTSD and DSO symptoms at T2, as well as with PTSD and DSO total scores at T1 and T2. Becoming hypersensitive to the perpetrator (IWA) was related to higher levels of affective dysregulation at T2 and DSO total scores at T1. Adopting the perpetrator’s experience concerning the abuse (IWA) was linked to higher levels of all DSO symptoms at T2, as well as with DSO total scores at T1 and T2. Identifying with the perpetrator's aggression (IWA) was associated with higher levels of all DSO symptoms at T2, as well as PTSD at T2 and total DSO scores at both T1 and T2.

Table 2.

Correlations between IWA, DARA, and symptoms of PTSD and DSO (n = 273).

Measure	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16
1. Replacing one's agency with that of the perpetrator (T1)	–
2. Becoming hyper-sensitive to the perpetrator (T1)	.42***	–
3. Adopting the perpetrator’s experience concerning the abuse (T1)	.58***	.57***	–
4. Identifying with the perpetrator's aggression (T1)	.45***	.45***	.69***	–
5. Doubt regarding the abuse (T1)	.19***	.19***	.35***	.28***	–
6. Doubt regarding the perpetrator (T1)	.18***	.24***	.28***	.26***	.53***	–
7. Doubt regarding oneself (T1)	.20***	.12***	.27***	.17***	.51***	.42***	–
8. Re-experiencing (T2)	.14**	.03	.03	.04	.04	.05	.17***	–
9. Avoidance (T2)	.12*	.06	.07	.08	.14**	.13*	.17**	.58***	–
10. Sense of threat (T2)	.12*	.04	.03	.10	.04	.08	.08	.53***	.62***	–
11. Affective dysregulation (T2)	.18**	.12*	.15**	.14*	.16**	.15**	.10	.45***	.58***	.59***	–
12. Negative self-concept (T2)	.12*	.03	.11*	.11*	.21***	.17**	.14**	.36***	.48***	.44***	.59***	–
13. Disturbed relationships (T2)	.18***	.07	.12*	.12*	.16**	.15**	.05	.43***	.58***	.52***	.70***	.58***	–
14. PTSD total score (T1)	.19**	.12	.11	.11	.08	.10	.29***	.44***	.35***	.31***	.19***	.21***	.19***	–
15. DSO total score (T1)	.25***	.14*	.21***	.26***	.23***	.18**	.27***	.22***	.27***	.22***	.32***	.41***	.35***	.31***	–
16. PTSD total score (T2)	.17**	.05	.05	.15***	.10	.12	.05	.74***	.80***	.79***	.61***	.47***	.56***	.39***	.25***	–
17. DSO total score (T2)	.20***	.09	.15*	.15*	.15**	.19**	.14*	.44***	.58***	.54***	.82***	.77***	.79***	.21***	.39***	.59***

p < .05. ** p < .01. *** p < .001.

Furthermore, doubt regarding the abuse (DARA) and doubt regarding the perpetrator (DARA) were both correlated with higher levels of avoidance and all DSO symptoms at T2 as well as DSO total scores at T1 and T2. Doubt regarding oneself (DARA) was linked to higher levels of re-experiencing, avoidance, and negative self-concept at T2, along with PTSD total scores at T1 and DSO total scores at T1 and T2.

3.3. The contribution of DARA and IWA at T1 in explaining PTSD and DSO symptoms at T2

Figure 1 presents the theoretical model, and Table 3 displays the results of the estimation process. We employed a hierarchical SEM, estimated in three steps: Model 1 included only the IWA and DARA predictors; Model 2 added polyvictimization; and Model 3 further introduced baseline PTSD and DSO symptoms at T1. Model fit improved at each step, with the most significant gain observed when baseline symptoms were included in Model 3 (CFI = 0.97, TLI = 0.95, SRMR = 0.04, RMSEA = 0.05 with CI = [0.032, 0.069]). The addition of polyvictimization in Model 2 had only a modest impact on the size and significance of the IWA and DARA coefficients. In contrast, the inclusion of T1 PTSD and DSO symptoms in Model 3 substantially reduced the magnitude of most structural paths from IWA and DARA to T2 outcomes, indicating that baseline symptom severity accounted for a considerable proportion of the variance in T2 PTSD and DSO. Nevertheless, the results also indicated that specific DARA and IWA components remained significant even after controlling for baseline symptoms, highlighting their independent predictive value for later outcomes.

Figure 1.

Theoretical model linking identification with the Aggressor, doubt regarding abuse-related abuse-related appraisals, polyvictimization and baseline symptoms (T1) to PTSD and DSO symptoms at follow-up (T2).

Table 3.

Results of Structural Equation Modeling (SEM) Predicting PTSD and DSO Symptoms at T2 Based on IWA and DARA at T1.

Measurement model
			Model 1		Model 2		Model 3
			$\hat{\beta }$(SE)		$\hat{\beta }$(SE)		$\hat{\beta }$(SE)
PTSD	Re-experiencing		0.701*** (0.035)		0.703*** (0.034)		0.718*** (0.033)
	Avoidance		0.854*** (0.029)		0.854*** (0.029)		0.834*** (0.03)
	Sense of threat		0.805*** (0.033)		0.804*** (0.033)		0.789*** (0.033)
DSO	Affective dysregulation		0.887*** (0.024)		0.887*** (0.024)		0.868*** (0.025)
	Negative self-concept		0.710*** (0.040)		0.710*** (0.040)		0.709*** (0.039)
	Disturbed relationships		0.863*** (0.025)		0.863*** (0.025)		0.861*** (0.025)
Structural model
		Model 1		Model 2		Model 3
		$\hat{\beta }$(SE)	$\hat{\beta }$(SE)	$\hat{\beta }$(SE)	$\hat{\beta }$(SE)	$\hat{\beta }$(SE)	$\hat{\beta }$(SE)
		PTSD	DSO	PTSD	DSO	PTSD	DSO
Adopting the perpetrator’s experience concerning the abuse		−0.206 (0.119)	−0.105 (0.108)	−0.222 (0.116)	−0.116 (0.107)	−0.168 (0.113)	−0.06 (0.100)
Identifying with the perpetrator's aggression		0.121
(0.093)	0.091 (0.088)	0.116 (0.089)	0.088 (0.087)	0.084 (0.089)	0.01 (0.083)
Replacing one's agency with that of the perpetrator		0.198* (0.077)	0.207** (0.074)	0.192* (0.076)	0.203** (0.073)	0.147* (0.070)	0.147* (0.067)
Becoming hyper-sensitive to the perpetrator		−0.001 (0.084)	−0.016 (0.078)	0.007 (0.084)	−0.01 (0.077)	−0.03 (0.076)	−0.021 (0.074)
Doubt regarding the abuse		0.007
(0.084)	0.159* (0.081)	0.054 (0.082)	0.193* (0.078)	0.071 (0.072)	0.17* (0.071)
Doubt regarding the perpetrator		0.072
(0.076)	0.102 (0.076)	0.06 (0.074)	0.093 (0.074)	0.064 (0.072)	0.084 (0.071)
Doubt regarding oneself		0.175* (0.073)	−0.015 (0.074)	0.152* (0.073)	−0.032 (0.073)	0.04
(0.070)	−0.097 (0.07)
Polyvictimization		–	–	0.195*** (0.060)	0.138* (0.062)	0.129* (0.058)	0.108 (0.059)
PTSD/ DSO at T1		–	–	–	–	0.407*** (0.053)	0.359*** (0.052)
$R^2$		0.08	0.10	0.12	0.11	0.27	0.23
Robust CFI		0.86		0.88		0.97
Robust TLI		0.82		0.82		0.95
Robust RMSEA		0.095 [0.081, 0.11]		0.10 [0.08, 0.11]		0.05 [0.03, 0.07]
SRMR		0.09		0.08		0.04
$\mathrm{\Delta }{\chi }^2$				9.78**		84.46***

***p < .001; **p < .01; *p < .05.

In the final model, T2 PTSD symptoms were significantly predicted by T1 PTSD symptoms ($\beta =0.407,p<.001$), the replacement of one’s agency with that of the perpetrator at T1 ($\beta =0.147,p=.035$), and polyvictimization ($\beta =0.129,p=.025$). T2 DSO symptoms were significantly predicted by T1 DSO symptoms ($\beta =0.359,p<.001$), the replacement of one’s agency with that of the perpetrator ($\beta =0.147,p=.028$), and doubt regarding the abuse ($\beta =0.17,p=.017$). Other predictors were not significant. Together, IWA, DARA, polyvictimization, and T1 PTSD and DSO accounted for 27.1% of the variance in PTSD symptoms and 23.3% of the variance in DSO symptoms at T2 (R² = 0.271, R² = 0.233).

4. Discussion

This two-wave study explored the relationship between DARA and IWA with later PTSD and DSO symptoms, as well as their predictive capacity over time, among female CA survivors. The results revealed significant associations between both DARA and IWA with symptoms of PTSD and DSO. Hierarchical SEM indicated that replacing one's agency with that of the perpetrator (a component of IWA), predicted subsequent PTSD and DSO symptoms, while doubt regarding the abuse (a component of DARA) predicted subsequent DSO symptoms. These effects remained significant even after accounting for polyvictimization, PTSD, and DSO symptoms.

Positive associations were found between the IWA total score at T1 and PTSD total scores at T1, as well as between the DSO total scores at T1 and T2. Additionally, the DARA total score at T1 was positively associated with PTSD and DSO total scores at both T1 and T2. The current findings align with prior research linking IWA to increased posttraumatic distress (Lahav, 2021a; Lahav, Talmon, & Ginzburg, 2019a; Lahav, Talmon, Ginzburg, & Spiegel, 2019b; Siegel et al., 2024), including DSO symptoms (Lahav, Cloitre, et al., 2025). They also support emerging evidence connecting DARA to PTSD symptomatology (Lahav et al., forthcoming; Lahav, Huberman, et al., 2025).

While IWA may help children cope during the abuse by predicting attacks, calming the perpetrator, and maintaining positive relations that may be crucial for the child’s well-being (Ferenczi, 1932, 1933; Lahav, Talmon, & Ginzburg, 2019a), its persistence after the abuse ends can have detrimental effects. These can include denial or minimization of the abuse, self-blame, self-directed aggression, and the repetition of abusive dynamics in later relationships (Lahav, Talmon, Ginzburg, & Spiegel, 2019b; Lahav et al., 2020). Moreover, as IWA is rooted in dissociative mechanisms, it hinders trauma processing and disrupts the formation of a stable, integrated belief system. IWA involves both concordant and complementary identifications (Frankel, 2002): survivors adopt the abuser’s perspective and aggression toward others (concordant) while also internalizing the abuser’s negative view of themselves as bad or blameworthy (complementary). This can lead survivors to view themselves, others, and relationships through a distorted, victim-perpetrator lens. Consequently, they may accept or anticipate abuse in close relationships, reenact abusive roles, and direct aggression both inward and outward (Lahav, Talmon, Ginzburg, & Spiegel, 2019b; Lahav et al., 2020, 2025). These patterns can obstruct trauma resolution, contributing to elevated concurrent PTSD symptoms and potentially facilitating the later development of DSO symptoms such as emotional dysregulation, negative self-concept, and relational difficulties.

Doubt Regarding Abuse-Related Appraisals (DARA) is a related yet distinct phenomenon that also appears maladaptive and is associated with heightened trauma-related symptoms. Survivors experiencing DARA struggle to confidently evaluate the nature of the abuse itself, their own behavior during the abuse, and the actions of their perpetrators. This persistent sense of doubt, which can continue long after the abuse has ended, significantly impairs survivors’ ability to construct a coherent and integrated narrative of the traumatic experience and its aftermath (Lahav et al., forthcoming; Lahav, Huberman, et al., 2025). Instead, it may fuel ongoing confusion, ambiguity, and psychological disorientation, leading survivors to oscillate between conflicting interpretations and undermining their trust in their own perceptions and judgments. Moreover, this experience can be particularly triggering, closely mirroring the psychological state during the abuse itself when survivors were overwhelmed, disoriented, and confused due to the perpetrator’s repeated attacks on their perceptions, as well as the activation of defensive reactions such as dissociation and IWA (Lahav et al., forthcoming; Lahav, Huberman, et al., 2025). Thus, DARA not only obstructs trauma processing, as reflected in posttraumatic stress symptomatology but may also contribute to unstable emotional responses, a diffuse and negative self-concept, and unsteady interpersonal orientations – all hallmark features of disturbances in self-organization (DSO) within the complex PTSD (CPTSD) framework.

The current hierarchical SEM indicated that while the total scores of IWA and DARA are related to PTSD and DSO, only specific components of these phenomena uniquely predict symptomatology in female CA survivors. Specifically, the aspect of IWA involving the replacement of one's agency with that of the perpetrator predicted both subsequent PTSD and DSO symptoms, while doubt regarding the abuse as part of DARA specifically predicted subsequent DSO symptoms. These effects exhibited stable and substantial explanatory power even after accounting for polyvictimization and baseline levels of PTSD and DSO symptoms. The predictive effect of relinquishing one’s agency aligns with emerging evidence highlighting this dynamic as a risk factor among survivors of child abuse (Lahav, Cloitre, et al., 2025). Importantly, this two-wave study extends prior research by demonstrating that this element of IWA not only correlates with psychological distress but also prospectively predicts trauma-related symptoms among female CA survivors. The tendency to replace one's agency with that of the perpetrator emerges as a particularly damaging aspect of IWA, contributing to both PTSD and DSO symptoms. Survivors who replace their agency with that of the perpetrator often lose touch with their own sense of self. Their subjective experiences become dissociated, and they grow disconnected from their emotions, needs, and desires (Frankel, 2002). Their sense of control over their actions is overtaken by the perpetrator’s will, resulting in deep mental subordination (Lahav, Talmon, & Ginzburg, 2019a). Even after the abuse ends, this dynamic–central to IWA–can impede trauma processing and severely disrupt basic functioning, as seen in complex PTSD (Lahav, Cloitre, et al., 2025). Psychological submission and disconnection from one's own emotions, impulses, and needs can damage self-concept, foster a sense of unworthiness or brokenness, impair emotional awareness and regulation, and undermine the ability to maintain healthy boundaries.

Simultaneously, the current findings revealed that while none of the DARA components predicted PTSD symptoms, doubt regarding the abuse specifically predicted subsequent DSO symptoms among female CA survivors. These results suggest that female survivors who face persistent doubt about the labeling, severity, and legitimacy of the abuse may experience heightened vulnerability to profound and enduring intrapersonal and interpersonal difficulties, as reflected in DSO symptoms.

Child abuse has long-term consequences that can profoundly shape the lives of survivors. Difficulties in making meaning of this central traumatic experience may therefore have far-reaching effects – extending beyond confusion or misperception to disrupt deep intrapersonal and interpersonal functioning. Female survivors with an incoherent or inconsistent narrative about their past abuse may struggle to understand and regulate their emotions. While they may suffer from the emotional and relational consequences of the abuse, ongoing doubt about its nature or legitimacy can obscure their ability to recognize these difficulties as trauma-related, thereby hindering appropriate coping or treatment. Moreover, as this sense of doubt can echo their past experiences of confusion and disorientation during the abuse, it may be experienced as a reenactment of the trauma itself – further complicating recovery over time. In this way, doubt regarding the abuse contributes to heightened DSO, as reflected in symptoms such as affect dysregulation, negative self-concept, and relational disturbances among the current sample.

The current study has several limitations. First, it relied on convenience sampling and utilized self-report measures, which may be subject to response biases and shared method variance. Second, other traumatic experiences were not assessed and therefore could not be controlled for. Third, potential mechanisms underlying the relationship between IWA, DARA, and PTSD and DSO symptoms – such as survivors’ current internalized subjugation or their attribution of difficulties – were not assessed. Additionally, several well-established correlates of DSO – such as attachment style, dissociation severity, ongoing interpersonal violence, socioeconomic instability, and personality-disorder traits (e.g. Karatzias et al., 2022; van Dijke et al., 2018) – were not measured, limiting our ability to rule out alternative explanations for the observed associations. While the findings suggest that IWA and DARA contribute to predicting DSO, the absence of these potential confounding variables warrants caution when interpreting the uniqueness of their effects. Furthermore, the current study did not examine the potential bidirectional relationships between the study variables over time and, therefore, cannot determine the extent to which PTSD and DSO symptoms may influence IWA and DARA. Although IWA and DARA are conceptualized as mechanisms underlying posttraumatic distress, existing scholarship suggests that these constructs may also emerge as consequences of symptom escalation. Specifically, difficulties in emotion regulation, self-worth, and interpersonal interactions – all components of DSO symptoms – may intensify survivors’ IWA and DARA. Although this bidirectional perspective cannot be empirically tested within the current design, acknowledging it situates our findings within the broader trauma literature and cautions against overinterpreting the observed associations as strictly causal.

The current sample consisted exclusively of adult Israeli women, so caution is warranted when generalizing these findings to male survivors or to women from other cultural and national contexts. Finally, although the sample size in this two-wave study was relatively small, which may limit statistical power and generalizability, it is noteworthy that several significant structural paths emerged despite this constraint. These findings suggest that the observed effects may be robust and meaningful, even under conservative conditions. To further advance our understanding of the implications of IWA and DARA for PTSD and CPTSD, future research should examine these variables over time using both self-report and clinical assessments, with larger and more diverse samples. Additionally, incorporating assessments of potential mechanisms and exploring the bidirectional associations between the variables will be essential for elucidating the processes underlying these associations.

Despite these limitations, the current study provides the first empirical evidence regarding the predictive effects of IWA and DARA on subsequent PTSD and DSO symptoms among female CA survivors, offering significant clinical insights. While the findings are preliminary, they suggest that IWA and DARA may serve as additional detrimental factors in complex trauma. Moreover, reducing both could potentially yield therapeutic benefits for individuals with CPTSD.

Although CPTSD is a relatively new diagnosis, there is emerging evidence for promising interventions for its treatment. These include modular therapies such as Skills Training in Affect and Interpersonal Regulation (STAIR) combined with narrative therapy (Niwa et al., 2022) or Enhanced Skills Training in Affect and Interpersonal Regulation (ESTAIR) (Karatzias et al., 2024). Addressing identification with the aggressor and doubt regarding abuse-related appraisals within modular treatment approaches for ICD-11 CPTSD can enhance their effectiveness. Specifically, the predictive role of relinquishing agency (a core IWA dynamic) and persistent cognitive doubt about the abuse (DARA) on PTSD and DSO symptoms suggests these constructs may represent key maintenance mechanisms of complex trauma symptomatology. Incorporating focused work on IWA and DARA within the Affective Regulation and Self-Concept modules could enhance treatment responsiveness by directly addressing dissociative submission, fragmented self-representations, and disrupted autobiographical coherence. Psychoeducation, cognitive restructuring, and narrative elaboration can help survivors recognize IWA and DARA as consequences of abuse that currently hinder their recovery. These approaches can also reduce identification with the perpetrator and bolster confidence in their own perceptions and interpretations of past abuse. Clinicians may explore how IWA and DARA relate to challenges in emotion regulation, self-concept, and interpersonal functioning, allowing for tailored interventions that address individual patterns. By directly addressing these processes, interventions can effectively reduce both IWA and DARA, facilitating the development of a clearer, safer, and more coherent autobiographical narrative. However, further research is needed to investigate the role of IWA and DARA in the treatment of CPTSD in child abuse survivors.

Funding Statement

This work was supported by The Nely Horovitz Research Fund for Neurological Diseases and The Walter Freundlich Foundation, Tel Aviv University .

Disclosure statement

No potential conflict of interest was reported by the authors.

Data availability statement

Due to the nature of this research, participants of this study did not agree for their data to be shared publicly, so supporting data is not available.